Why You Shouldn’t Care About Digital Clipping

I see a lot of talk on the interwebs about digital clipping and how it’s a mortal sin, you know how it goes “analog clipping is good, and digital clipping is bad”.

I feel like this is a spread of misinformation.

In theory, yes, digital clipping is “bad” but in practice, it’s not that big of a deal. In fact, it can be a REALLY good thing. It allows you to push audio beyond its limits to get something you would not have been able to otherwise.

But before I show you why you shouldn’t care about clipping that much, let’s try and get an understanding of what it is.

What is Digital Clipping?

To not bore you with some long-winded technical spiel, I will give you the kindergarten version. Okay, maybe it’s not just to bore you but because I’m not a very technical engineer; everything I do is from feel and experience, so I’m not exactly the best person to explain this from a technical standpoint.

More technical knowledge over at Wikipedia

onto to what digital clipping is…

There’s only so much space that your audio files have.  You can keep turning the level up but at some point we are going to reach our maximum level and then this is where clipping would occur and the audio signal craps out (nasty digital distortion).

At that point, our nice round peaks that we worked so hard to achieve during recording start to square off.

clipping2

As you can see from my (less than average) drawing above, the left side represents a normal sine wave, and the right side represents a clipped sine wave.

The red line at the top and bottom represent the maximum level that the wave can reach before clipping occurs (0dBFS).  You can see that when a digital wave clips it is completely chopped (squared) off and isn’t forgiving at all.

In the analog world, clipping is a little bit more forgiving and tends to round the edges off. This means that the sound will be a lot less abrasive and in most cases, it does something nice to the sound.

Digital clipping is like slamming into a brick wall where analog clipping is more like having a bunch of mattresses in front of the wall to soften the blow.  Not the best analogy but you get the idea.

Gain Staging and Headroom

I’m a huge proponent of gain staging, which I practice on every mix and encourage everyone I teach to do the same. That said it isn’t uncommon to see some clipping in just about every mix that I do which you will see in my tutorials.

Digital mixing and recording have come a long way since the 16 bit days so clipping is almost irrelevant…almost!

In my opinion the only time you really have to be strict about headroom (when your mixing a song) is in 2 situations:

#1 When You Are Going Into Plugins

If you aren’t familiar with your plugins input and output stage you could be adding in some nasty digital distortion.  Although a lot of plugins now come equipped with over sampling and also analogue style output stages (pleasing distortion), it’s still good practice to leave a fair amount of headroom if you can.

#2 Coming Out of Your DAW and Going into Your Convertors

 You almost never want to leave your DAW (via digital to analog conversion) while it’s slammed in that reds.  I just can’t think of a situation where that would be desirable as it leaves you vulnerable to making terrible mixing decisions based on the distortion you’re hearing.

Clipping Ain’t So Bad

clipping2With all that said, clipping inside your DAW is actually not a bad thing thanks to 32 bit and 64 bit floating point processing. With floating point processing, even though the clipping light is red, the audio file is actually not clipping.

Floating point processing essentially gives you an infinite amount of headroom, as long as the audio stays inside the digital realm.  That means it’s impossible to clip your audio files inside Pro Tools (or whatever DAW you use). But if you aren’t careful, once the “clipped” audio leaves your DAW, you will hear the nasty effects of digital distortion.

If you aren’t sure if your DAW is equipped with 32 bit or 64 bit floating point, I wouldn’t worry as I haven’t seen one modern DAW without that capability.

Also there won’t be any need to convert your 24 bit files as your DAW will do everything for you internally and then convert it back at the output stage.

So next time you see one of your fader meters light up red, try to think of it as a warning that your signal has gone over 0dBFS and not that it is distorting – use your ears for that.  I am strictly talking about the channel faders and not the master fader of course.  If your master fader is lighting up red then that will be a problem for your convertors and your mixes.

More on audio bit depth here.

Also, I do use clipping as part of my “mastering process”, which is separate from what I do while mixing.  If you want to see that technique and how I apply it to mastering, please click here.

So How Should I Approach Digital Clipping Then?

The first thing you should do is eliminate the thought that clipping is pure evil because it isn’t. Digital clipping is kind of like getting sick (not terminally), you want to avoid as much as possible but when it happens it might be a good thing – i.e., Time off work or maybe a reason to avoid the in-laws ;).

Here’s a quick statement on what Phil Tan thinks about clipping.

A little red here or there – That’s not going to hurt nobody.  If you start to hear, things crap out then that’s a bad thing.

The quote starts at about 47:35 and if you scroll back from there a bit you will seem some of his faders clipping.

[youtube id=”kPH9LrLAmmo?t” width=”700″ height=”394″ autoplay=”no” api_params=””]

When Phil Tan is working on a mix, do you think he sits there and stresses over ever peak in his session?  Probably not.

I think it’s safe to say that he just focuses on making the record sound good.  If something does clip than it only becomes an issue if it affects the sound negatively.

Some Reasons Why My Meters Clip

I don’t set out to clip my faders but when it does happen, I’m not usually bothered by it.

I find that when I’m clipping it’s either from some sort of parallel compression technique or because I am really trying to drive a plugin’s input or output.

When I am using parallel compression, I am usually pushing the compressor to its limits and it’s not uncommon for it to start hitting the reds. A lot of times I won’t adjust the output because the plugins output is meant to distort (analog style) and I want that colour added to my track(s).

Another time is when I drive a plugin’s input stage (like the HoRNet SW34EQ) to give it some of that Analog color, it tends to push the output up into the reds. Again I would normally just leave it and only adjust if I feel like it is adding something negative to the sound. If another plugin is coming after then I would either adjust the input of the new plugin, so it’s not clipping, or I would adjust the output of the plugin that is causing the clipping. It ultimately just comes down to whichever method sounds and works the best.

By the time it all gets mixed together, you shouldn’t see any red lights on your mix buss, if you do that is probably a good indication that you need to start adjusting some faders to compensate for that.

The Take Away

Remember digital clipping only becomes a problem with certain plugins and when you are leaving the digital world.  It’s definitely better to err on the side of caution but if you see the odd red light here or there, that’s okay.

The best thing to do is listen to the track in question.  If you hear unpleasing digital distortion being added to your track, than correct it.  If you don’t, just leave it.

 

  • Valter

    Trust your untrained ears and a set of mediocre speakers in an untreated room you say? So what happens when those mixes hit a few kW of piezo horns? And what happens when they hit a broadcasting mixer? Standards are there for a reason. Only when you fully understand the rules you can start bending them. You might, but the kids reading this don’t.

  • “The quote starts at about 47:35 and if you scroll back from there a bit you will seem some of his faders clipping.”
    And when you listen how it sounds there…it’s horrible! The kick is distorted and everything sounds crap. So what’s your point? Maybe they clipped in the recording of the video? So is that something to care about or not?

    I understand what you say overall in the article but it is just terribly put down. The “why you shouldn’t care about digital clipping” is just to gain clics man and you know it as much as I do.

    You have to care about clipping that is why you decide to do it or not, and also you need to know what your meters mean as they might be red but you are not necessarily clipping because the calculation of the signal might be in floating points.

    You might be good at mixing and working sound but you should pay a little more attention on your writing skills and how to convey information.

  • John Smith

    It’s because of clipping and over compression that modern day music can’t be listened to for more than a couple of minutes before your ears scream at you to turn it either down or off. Off course, because it’s so over compressed, when you turn the volume down, the song then sounds completely dill, flat and lifeless. Music is supposed to have varying peak lengths. Sometimes the drum is supposed to sound harder and louder and sometimes it’s supposed to sound softer and quieter. It’s the variation in the dynamics that makes the music exciting and enjoyable to listen to. Don’t ruin your music by taking that away from it.

  • Samuel

    ok so my issue is I cant get my tracks to sound loud enough without making them clip or producing them at ridiculous high levels.. is it chill to produce your track clipping in the red (master) then at the end just throw a limiter on the master? then go to mastering? this tune im workin on now sounds dope as shit without anything on the limiter but its clipping +8 on the master lol .. this was an exp track but now its one of my better ones! ha. I just know im committing audio engineer sin right now. If someone could please help thatd be great. its not mixing issues either. Its just i like my kicks/808 loud but i can’t figure out how to make them loud and still not clip. i do lots of rev. eq’n when needed but i tend to just side chain my sub to kick. i dno if im making sense i just need some haaalp! thanks yall

    • Oops

      I know this is a couple weeks old but if you’re still looking at this, tell me am I right in assuming you’re mixing on headphones? Or computer speakers even?

    • In terms of what I do? I’ll generally keep things away from the reds if at all possible but on the final pass/mix down. I’ll usually run the entire mix through my convertors to where it clips. It gels the entire track together and give it a “sound” that I like. It doesn’t work on everything but when it does work, it works like crazy. You should have REALLY good convertors to make it work though.

  • darknesscrown

    As a professional engineer who HAS been formally trained in this, I really need to say that young “producers” in their bedrooms and basements are going to get the wrong idea here. Digital clipping going into the A/D converter is ALWAYS bad. This is not an opinion. This is math. The converters assign a binary word, rounded to the least significant integer, to every sample of electric voltage hitting the converter on the input…and this process ALWAYS results in imperfections during quantization to binary words. This quantization error CANNOT be eliminated.

    The MOST IMPORTANT part of conversion is A/D…NOT D/A. If you clip, the A/D converter as you are capturing audio, you are screwed. Period. No argument. No debate. You mentioned gain staging, that is absolutely critical…and it, too, is not something that is an “artistic liberty” thing. There is a precise procedure for setting your gain stage. It depends on many factors such as whether or not the audio source is active or passive and performing proper impedance matching. Things like that. If you do not understand any of that, you will soon enough in this business…because if you don’t learn it you won’t be doing this kind of work very long, that’s for sure.

    With regard to 32 (or 64) bit float points, I’m going to tell you right now, don’t even bother setting your session to that. It isn’t doing anything. it doesn’t give you more “headroom” in the digital domain, at all. If your system is 24 bits, then that’s all it is. It is absolutely a waste of time to set your session to 32 bit float unless you have (or have access) to a system that processes audio at 32 bits. And even then, if you are clipping in 32 bit, you’re done.

    Bottom line here is this: EXCEEDING 0dBFS on your meters in your DAW is ALWAYS bad. Period. Especially if you do it while tracking…if it happens then, you’re fucked. The audio is useless and CANNOT be fixed. Ever. It’s less of a catastrophe coming OUT of the digital domain, but you should NEVER be clipping then, either. So the lesson here is ignore the idea that “[digital] clipping can be a good thing” and just don’t do it. Analog is not clipping until you HEAR distortion occur – it is much more forgiving. If your DIGITAL meters are clipping, you’ve fucked it up.

    • You are wrong on a couple points. You should definitely mix with your ears and not your eyes. You got it wrong.

    • Jeremy Joel Signorelli

      Even though sound is physics, and physics is directly linked to math in regards to your post, it is in my belief that you have to mix with your ears in regards to the kind of music or effect you’re going for. Plenty of modern pop mixes are in the red almost all the time if you bring it into a DAW and look at it. It almost makes me cringe. But when I listen to some of them, they don’t sound like they’re clipping. They may a little, but it’s not the “crappy” kind of clipping. Sometimes, to me, it gives a certain dirty grunge that is appealing, such as Charlie Puth’s new mix, “Does It Feel”. It is so loud, has so much punch, and is almost constantly in the red. I suppose it just depends on the sound you want in the end.

    • Craig

      You seem to misunderstand 32-bit floating math. Bottom line: your DAW has 1500dB of extra headroom via Floating Point math. ALWAYS mix in “float” if your DAW has it. I also don’t drive any of my tracks into the red. However, you could take 10 tracks and run them at +12 — screaming bloody red, and just pull down your master to compensate: and no clipping. Try it!
      But you’re absolutely right about recording levels. Record 24-bit tracks aiming for average around -20dB with peaks around -12 and you’re in great shape. NEVER clip on input!!!

    • Rob

      I understand what you’re saying, but tracking with a digital distortion is different than mixing VST’s and pre-recorded audio (with digital distortion). If you’re in the red a bit with a sampled kick drum, it’s really not that big of a deal. Especially if it still sounds good. However, if you’re recording a vocal and clipping, that’s a totally different issue.

  • Motherfuckin’ Charlie Bronson

    I kept hitting the red line, and kept thinking “this really sounds really good”, but then remembered what my grandfather said about clipping, and how it was the work of Satan. Glad I read your article; I’m going to trust my ears more.

    • Glad to have you here! As a best practice I would avoid the reds whenever possible but if driving the output of a plugin into the reds on a kick drum (lets say) sounds good, then it is good. Cheers!

  • Chris Wellz

    Very good read! I remember reading this a while ago. Glad I found it again.

  • Jerry

    Even Pensado digitally clips his tracks once in a while, so I guess it’s ok, maybe 🙂

  • Chuck

    I see your point, but why clip the faders (and potentially undermine the algorithms) when you can just a digital clipper plugin designed to handle signal in that way? With this approach there is no reason for any red ever, and I can tell you from practice it results in a cleaner sound while providing the desired overloaded effect. The fader is not an effect, so don’t use it like one.

  • Okay, so basically you’re saying: clipping on your channel strip faders isn’t a bad thing – but never clip your master bus fader.

    Interesting and true, but if a channel strip fader is clipping, won’t that translate to your master bus fader (sitting at 0db) clipping? OK to lower the volume of the master fader in that case, but I actually rather make sure it doesn’t clip set at 0db.

    I’ll have to check that when I get to the studio tonight :-).

    Thanks for the interesting take on this!

    Cheers, Kevin

    • Justin Smith

      not if your meters are set pre fader. Cheers.

      • Got a question, I am recording a song from a free-beat found on youtube. and the actual beat comes into the DAW with hitting the red zone, even after a mix.. my vocals I recorded do not go into red zone but the actual beat makes parts of the entire track go into red.. how do I fix this, if possible?

        • You can’t fix a beat that’s been squashed. What you can do is lower the gain input on the beat so that it’s not slamming up at 0dBfs. take it down to like -10dBfs and then mix your vocals into that. Once you are good to go, that you can raise the volume of everything together.

Still Struggling with EQ?